作者: Matthieu Chartier , Rafael Najmanovich
关键词: Similarity (network science) 、 Receiver operating characteristic 、 Function (mathematics) 、 Data mining 、 Drug toxicity 、 Interaction field 、 Protein family 、 Molecular recognition 、 Biology 、 Computational biology 、 Binding site
摘要: Protein binding-site similarity detection methods can be used to predict protein function and understand molecular recognition, as a tool in drug design for repurposing polypharmacology, the prediction of determinants toxicity. Here, we present IsoMIF, method able identify binding site interaction field similarities across families. IsoMIF utilizes six chemical probes subgraph isomorphisms geometrically chemically equivalent sections cavity pairs. The is validated using distinct data sets, four those previously validation other methods. mean area under receiver operator curve (AUC) obtained sets higher than Furthermore, while obtains consistently high AUC values perform more erratically sets. from structure, detect poten...